A repository dedicated to recreating popular machine learning algorithms and models from scratch in python. It is a mental exercise to deepen understanding through practical examples.
Out-of-the-box machine learning algorithms are available everywhere. Most people know when and how to use these algorithms, but very few know the intricate workings behind them. I took a class on Multivariate Statistics that explored these algorithms from a theoretical perspective. I would like to continue this exploration in code. In this repository, I will be implementing these algorithms from scratch. The motivation is to improve coding skills, better understand algorithms, and learn how to implement them across different datasets. I built this project as a guide for someone looking to improve their machine learning skills alongside me.
I will be working mainly on the algorithms and implementation through theory and knowledge. I expect to make mistakes, but I also expect to get better as I go.
The first algorithm I will explore is K-nearestneighbor (KNN). It is a pretty simple algorithm to implement and quite strong out of the box.